A ruling relevant to free software was recently made by a judge presiding over the highly publicized trial of Kyle Rittenhouse in Kenosha, Wisconsin, US.
On the night of August 25 2020, Kyle Rittenhouse shot three men with a rifle in Kenosha, Wisconsin during heated protests following a police shooting of a black man. Rittenhouse, a resident of Antioch, Illinois was 17 at the time and by law, prohibited from owning a gun or porting one across state borders. Nobody had invited him to Kenosha. He claims innocent on the charges brought to him. The fatal shootings, according to his lawyers, was an act of self-defense. For further information on the incident and the trial, please read this Wikipedia article: Kenosha unrest shooting - Wikipedia https://en.wikipedia.org/wiki/Kyle_Rittenhouse We know that Rittenhouse shot his rifle at point blank during confrontations with his victims. The cause of these confrontations is unclear. Was there provocation? If so, from which side? The prosecutor has supplied video footage in their attempt to clarify. However the videos, taken at night, poorly focused are hard to discern. At one point the prosecutor tried to show the jury an enlarged image as evidence, but was turned down on the grounds that the process generates pixels which do not exist in the original and among these additional pixels there may be some that comprise an illusion that jurors may interpret as incriminating. Rittenhouse's lawyers argue that zooming in on a video could distort the image. https://www.nytimes.com/2021/11/10/us/kyle-rittenhouse-video-zoom.html Blow-up at Rittenhouse trial over enlarging photos and video https://www.washingtonpost.com/national/blow-up-at-rittenhouse-trial-over-enlarging-photos-and-video/2021/11/11/ebc371a4-4354-11ec-9404-50a28a88b9cd_story.html --- I work on image processing software. I may, one day, be summoned to court to testify on the technology in use here. Software that enlarges images add pixels, but generally this is done by interpolation. A new pixel between an existing red pixel and a yellow pixel is colored orange. By this method, high-contrast details do not emerge from the void. But if the lawyers on the opposing side decide to argue that advanced AI is capable of more sophisticated things, I would not be able to offer much of a counterargument. Without access to the source code, no third-party observer can say for certain. This is an example of what proprietary software does to your data. Here we have a lawyer successfully arguing that potential distortions caused by undisclosed technical features makes a video unacceptable as evidence. We should encourage manufacturers of popular portable devices to consider additional functions which would ensure that audio and video recordings will stand up in court. The algorithms should be kept simple and, more importantly, be fully disclosed.